3. The decision to return to school

View the most recent version.

3.1 Basic framework

Dropping out is largely characterized in the economics literature as the result of a rational decision.¹⁰ Individuals consider the benefits and costs of continuing school (i.e. to continue and complete their high school diploma versus to leave school). This rational decision is driven by low academic ability (lower benefits), low expectations about the rewards of further education (lower benefits), and appealing local work opportunities (higher opportunity costs). Eckstein and Wolpin (1999) identify two types of dropouts: the academically weak who drop out early on and those with low expectations about the rewards from graduation who may drop out as late as in their last year of high school.¹¹ They argue that dropouts have a comparative advantage – the type of abilities and interests, for example manual skills or interest in physical work – for the jobs they obtain with the level of schooling they have. It implies that young adults decide to drop out of school to take up employment that requires the type of skills that they have (without further education).These decisions are based on local labour market conditions, including the minimum wage and the work opportunities accessible at the time to young decision makers (Eckstein and Wolpin, 1999; Chaplin et al., 2003; in Canada, Parent, 2006). Some recent evidence suggests, however, that high school students' decisions have become less sensitive to work opportunities and more to wage premiums over the 1990s (Ferrer and Lauzon, 2005).

All the previously-cited studies also find some influence of personal and family backgrounds on the decision to drop out. Dropouts are more likely to come from a single-parent household, from a low-income family and/or from a family where the parent does not have a postsecondary diploma or degree. Oreopoulos et al (2003) show that parental education captures more than just parental ability, which could be inherited by children, by using compulsory schooling laws as instrument. Their findings imply that independently of ability, educated parents stress the importance of education to their children. With a lower appreciation of schooling, dropouts exhibit lower schooling aspirations and are more likely to engage in delinquent behaviour like drinking (descriptive work in Canada, Bowlby and McMullen, 2002; Chatterji and DeSimone, 2005). It is furthermore noteworthy that future dropouts exhibit these characteristics as early as age 15 and several months prior to leaving school (Bushnik et al., 2004).

The above rational decision framework implicitly assumes perfect information and perfect credit markets. Lack of information may lead to an erroneous valuation of the net benefits of graduating from high school, while borrowing constraints may force dropping out to work. As individuals acquire new information about the benefits and costs of schooling, they may reconsider their decision and decide to return to school. Similarly, individuals may build their credit or savings to relax borrowing constraints. The Becker model of human capital investment can be modified to include such schooling pathways by incorporating imperfect information and borrowing constraints (Becker, 1962).¹² Upon the acquisition of new information or loosening of borrowing constraints, individuals can reconsider their past decision and may reverse it. Altonji (1993) uses this augmented model to explain change in field of study or occupational choice and Light (1995), to explain schooling interruptions between high school graduation and postsecondary schooling. Various studies have investigated the effect of borrowing constraints – referring to the case where the individual does not have the financial resources to study and does not have access to sufficient credit, whether a governmental loan or private credit from a bank – on postsecondary educational attainment (see for example, Keane and Wolpin, 2001).

In a model which incorporates incomplete information and borrowing constraints, individuals may leave school without graduating and may later return to school based on new information or new financial resources. The present analysis only focuses on the process of re-evaluating schooling net benefits and returning to school, not the previous decision of dropping out.

Individuals acquire new information and use it to update their evaluation of the net benefits from schooling. New information may pertain to the earnings a high school dropout gives up to attend school (opportunity cost), the earnings advantage a high school graduate has (benefits), the relative instability of employment for a dropout versus a graduate (expected benefits), or preferences about work and school (benefits). Besides these elements, the individual will also factor into his decision attendance costs such as school materials and tuition. The costs are scaled to what remains to be completed, in other words, whether only grade 12 needs to be completed or grades 10 through 12.¹³ The costs have to be weighed against financial resources and needs, including new ones, such as living arrangements and dependents.

Other types of constraints may lead to dropping out. Health problems, family problems, personal problems may all make it almost impossible to attend school. Whether facing borrowing constraints or personal problems, the individual may be convinced of the value of schooling and may even aspire to pursue a postsecondary education. She or he may choose to temporarily leave school and plan to return at a later date once the constraints or barriers have eased up. Such temporary leavers would be more likely to return than individuals who have to re-assess the net benefits. The economic research has not paid much attention to these intentionally-temporary dropouts (or stopouts).

Only a few economists have investigated returns or re-enrolments. In the U.S., Light (1995) looks at high school graduates who interrupt their studies before pursuing some postsecondary schooling. A few studies have analyzed schooling interruptions during postsecondary studies, referred to as stopouts, mostly at the university level (see for example, Singwell, 2001 and Stratton et al., 2005). Returns by high school dropouts have received very little attention. In Canada, Bushnik et al (2004) portray returnees as being equally likely to be male or female, more likely to be from Quebec and less likely from Alberta, less likely to be working full-time, and more likely to have parents with a postsecondary diploma. Interestingly, the study also reports that dropouts with postsecondary education aspirations, directly measured by asking what highest diploma or degree they would like to obtain, were more likely to return. Finally, only Chuang has empirically investigated reenrolment of dropouts for the U.S. (1994, 1997). He finds that ability as measured by a standardized test, age, the duration of out-of-school period, and local labour market conditions influence the decision to return. Neither family background nor the dropouts' activities during their out-of-school period have much explanatory power in predicting who returns. No study has attempted to explore the phenomenon of intentionally-temporary dropouts and whether returns are more common among them.

3.2 Two distinct groups of decision makers?

As noted above, young men drop out more and are less likely to return to school than young women in Canada. Many studies ignore these gendered-differentiated patterns and the few that note them do not attempt to explain the causes of these differences. For example, Eckstein and Wolpin (1999), Chuang (1994), and Light (1997) focus on men. Ferrer and Lauzon (2005) include a sex dummy variable and note that the gender gap appears to close over time. Parent (2006) conducts a separate analysis by gender but does not offer any explanation for the observed differences.

Yet, quite a few stylized facts have been highlighted in the literature. Parent's work (2001) reveals that women's decision to drop out appears less sensitive to labour conditions or to having worked during high school than men's. Having a child has a greater influence on women's decisions than on men's. Work by Bowlby and McMullen (2002) echoes these findings. Young men cited wanting to work as a reason to leave school more often than young women. Furthermore, male dropouts appear to have a lower academic record, including lower grades and a higher incidence of grade repetition, than their female counterparts (Bushnik et al., 2004). American studies found similar patterns (see Chuang, 1997).

The above evidence suggests that male and female dropouts constitute two distinct groups and most likely, different categories of potential returnees. Men and women experience different circumstances leading to dropping out and appear to differ in preferences for school and work. One can speculate that the proportion of intentionally-temporary leavers varies across gender and the difference may partially explain the gender gap in returns to school. The remainder of the gap would be the result of differentiated preferences for school and work.